Abstract
Sample size calculations are an important part of research to balance the use of resources and to avoid undue harm to participants. Effect sizes are an integral part of these calculations and meaningful values are often unknown to the researcher. General recommendations for effect sizes have been proposed for several commonly used statistical procedures. For the analysis of 2×2 tables, recommendations have been given for the correlation coefficient φ for binary data; however, it is well known that φ suffers from poor statistical properties. The odds ratio is not problematic, although recommendations based on objective reasoning do not exist. This paper proposes odds ratio recommendations that are anchored to φ for fixed marginal probabilities. It will further be demonstrated that the marginal assumptions can be relaxed resulting in more general results.
| Original language | English (US) |
|---|---|
| Article number | e58777 |
| Journal | PLoS One |
| Volume | 8 |
| Issue number | 3 |
| DOIs | |
| State | Published - Mar 7 2013 |
| Externally published | Yes |
Fingerprint
ASJC Scopus subject areas
- Agricultural and Biological Sciences(all)
- Biochemistry, Genetics and Molecular Biology(all)
- Medicine(all)
Cite this
Effect Sizes for 2×2 Contingency Tables. / Olivier, Jake; Bell, Melanie L.
In: PLoS One, Vol. 8, No. 3, e58777, 07.03.2013.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Effect Sizes for 2×2 Contingency Tables
AU - Olivier, Jake
AU - Bell, Melanie L
PY - 2013/3/7
Y1 - 2013/3/7
N2 - Sample size calculations are an important part of research to balance the use of resources and to avoid undue harm to participants. Effect sizes are an integral part of these calculations and meaningful values are often unknown to the researcher. General recommendations for effect sizes have been proposed for several commonly used statistical procedures. For the analysis of 2×2 tables, recommendations have been given for the correlation coefficient φ for binary data; however, it is well known that φ suffers from poor statistical properties. The odds ratio is not problematic, although recommendations based on objective reasoning do not exist. This paper proposes odds ratio recommendations that are anchored to φ for fixed marginal probabilities. It will further be demonstrated that the marginal assumptions can be relaxed resulting in more general results.
AB - Sample size calculations are an important part of research to balance the use of resources and to avoid undue harm to participants. Effect sizes are an integral part of these calculations and meaningful values are often unknown to the researcher. General recommendations for effect sizes have been proposed for several commonly used statistical procedures. For the analysis of 2×2 tables, recommendations have been given for the correlation coefficient φ for binary data; however, it is well known that φ suffers from poor statistical properties. The odds ratio is not problematic, although recommendations based on objective reasoning do not exist. This paper proposes odds ratio recommendations that are anchored to φ for fixed marginal probabilities. It will further be demonstrated that the marginal assumptions can be relaxed resulting in more general results.
UR - http://www.scopus.com/inward/record.url?scp=84874703407&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84874703407&partnerID=8YFLogxK
U2 - 10.1371/journal.pone.0058777
DO - 10.1371/journal.pone.0058777
M3 - Article
C2 - 23505560
AN - SCOPUS:84874703407
VL - 8
JO - PLoS One
JF - PLoS One
SN - 1932-6203
IS - 3
M1 - e58777
ER -